In many small businesses, the manager's day works like a queue. A quote is waiting for a go-ahead, a supplier asks whether the delivery date can move, an employee wants to know if they can give a good customer a 5% discount. Each question is small. Added together, they turn the manager into the company's bottleneck, and a week of their absence into a week of slow motion.
The problem is usually not a lack of trust in the team. It is that nobody ever wrote down who can decide what. When in doubt, people ask. And since the manager always answers, they keep asking.
Why "delegate more" isn't enough
Telling the team to "take more initiative" changes nothing while the boundary stays vague. Someone who was pulled up once for deciding alone won't take that risk again. What unblocks things is an explicit rule: for this type of decision, up to this amount or within this frame, this person decides, without asking.
The table: one row per recurring decision
Take a sheet of paper or a spreadsheet. For two weeks, note every question you are asked that needs a decision. You will quickly see that they are always the same ones, fifteen or twenty of them. For each, fill in four columns:
- the decision, stated concretely: "give a discount", "order stock", "accept a new customer on credit terms";
- who decides: one named person, not "the team";
- the limit: an amount, a time frame, a condition ("up to €500", "if the customer has no overdue invoice");
- who needs to be told afterwards, and how: a message, a note in the file, or nobody.
Consultants will recognise a stripped-down RACI matrix (responsible, accountable, consulted, informed). For a small business, four columns are enough. What matters is that the "who decides" column only ever holds one name. Two names guarantee that each will wait for the other.
Setting limits without scaring yourself
Start with cautious limits that you can raise later. A discount up to a certain percentage, an order up to a certain amount, a goodwill gesture with a cap. Beyond that, the person comes to you, but with a proposal, not an open question. "I suggest 8% because this customer orders three times a year" takes ten seconds to settle. "What should we do about this customer?" takes a quarter of an hour.
Keep for yourself what really commits the business: hiring, signing a long contract, changing a list price, anything involving the bank. That is not mistrust, it is the normal split of roles.
Making the table last
The table is only worth something if both sides stick to it. For the team, that means deciding within the frame without asking "just to be sure". For the manager, it is harder: when someone brings you a decision that is in their column, hand it back. "That's your call, the table says so." The first two or three times cost a little. After that, the questions stop.
If a decision made within the frame doesn't suit you, don't overturn it in front of everyone. Adjust the limit or the rule in the table, and explain why. Overruling an authorised decision cancels the whole table.
Review it every three months at a team meeting: what is missing, what got stuck, which limits can go up. A table that is kept alive stays useful; one filed away in a shared folder is forgotten within six weeks.
Where to start this week
Don't try to cover everything. Take the three decisions you are asked about most often, write the rule for each, and announce it to the team. If you have already mapped a process, the points where work waits for approval are exactly the rows to put in the table. And protect the time you get back: it is the moment to restart uninterrupted blocks of work.
If you'd like to do this exercise with an outside view, especially while a team reorganisation is under way, it is the kind of work we do through hour packs: info@z80.be or +32 494 111 041.